ORIGINAL PAPER

Zanzibar and Indian Ocean trade in the first millenniumCE: the glass bead evidence

Marilee Wood1& Serena Panighello2,3 & Emilio F. Orsega2 & Peter Robertshaw4

&

Johannes T. van Elteren3& Alison Crowther5 & Mark Horton6

& Nicole Boivin7

Received: 15 June 2015 /Accepted: 22 December 2015 /Published online: 20 January 2016# The Author(s) 2016. This article is published with open access at Springerlink.com

Abstract Recent archaeological excavations at the seventh-to tenth-century CE sites of Unguja Ukuu and Fukuchani onZanzibar Island have produced large numbers of glass beadsthat shed new light on the island’s early interactions with thewider Indian Ocean world. A selected sample of the beadsrecovered was analyzed by laser ablation-inductively coupledplasma-mass spectrometry (LA-ICP-MS) to determine the or-igins of the glass used to make the beads and potential traderelationships are considered. The data show that two major

glass types can be identified: mineral-soda glass, m-Na-Al,produced in Sri Lanka (and possibly South India) and plantash soda glass. The latter comprises three subtypes: two withlow alumina concentrations and different quantities of lime(here designated v-Na-Ca subtypes A and B) and one withhigh alumina (designated v-Na-Al). The v-Na-Ca subtype Abeads are chemically similar to Sasanian type 1 glass as wellas Zhizo beads found in southern Africa, while v-Na-Ca sub-type B compares reasonably well with glasses from Syria and

Marilee Wood is a research associate at the University of theWitwatersrand and can be conctacted at the provided email address.

Electronic supplementary material The online version of this article(doi:10.1007/s12520-015-0310-z) contains supplementary material,which is available to authorized users.

* Marilee Woodmwood@rockisland.com

Serena Panighelloserena.panighello@unive.it

Emilio F. Orsegaef@unive.it

Peter Robertshawproberts@csusb.edu

Alison Crowthera.crowther@uq.edu.au

Mark HortonMark.Horton@bristol.ac.uk

Nicole BoivinNicole.boivin@riaha.ox.ac.uk

1 School of Geography, Archaeology and Environmental Studies,University of the Witwatersrand, Johannesburg, South Africa

2 Department of Molecular Sciences and Nanosystems, UniversityCa’Foscari of Venice, Calle Larga S. Marta 2137,30123 Venezia, Italy

3 Laboratory for Analytical Chemistry, National Institute of Chemistry,Hajdrihova 19, SI-1000 Ljubljana, Slovenia

4 Department of Anthropology, California State University, 5500University Parkway, San Bernardino, CA 92407-2397, USA

5 School of Social Science, The University of Queensland,St Lucia 4072, Australia

6 Department of Archaeology and Anthropology, University of Bristol,43 Woodland Road, Bristol BS8 1UU, UK

7 Research Laboratory for Archaeology and the History of Art,University of Oxford, Dyson Perrins Building, South Parks Road,Oxford OX1 3QY, UK

Archaeol Anthropol Sci (2017) 9:879–901DOI 10.1007/s12520-015-0310-z

the Levant. While the mineral-soda beads were made in SouthAsia, it appears likely that at least some of the plant ash beadswere made in South or Southeast Asia from imported raw and/or scrapMiddle Eastern glass. In contrast, during this period, allbeads imported into southern Africa were made of MiddleEastern glass from east of the Euphrates (v-Na-Ca subtype A)and appear to have arrived on ships from Oman and the PersianGulf. These data suggest that the two sections of the Africancoast were engaged in different Indian Ocean trade circuits.

Keywords LA-ICP-MS . Archaeological science . EasternAfrica . Middle East . South Asia . Southeast Asia

Introduction

In 2011 and 2012, the Sealinks Project (www.sealinksproject.com/) undertook excavations at two archaeological sites onZanzibar that possess evidence for trade contacts with thewider Indian Ocean world in the second half of the firstmillennium CE. The small site of Fukuchani lies on thenorthwest coast of the island, and Unguja Ukuu, which was asignificant port town, lies on the southwest coast more or lessopposite present-day Dar es Salaam (see Figs. 1 and 2 for mapsof the Indian Ocean and Zanzibar). Earlier excavations at thetwo sites (Horton and Clark 1985; Horton andMiddleton 2000;Horton 2015; Juma 2004) provided evidence of Indian Oceantrade, but the Sealinks Project excavations were undertakenwith the aim of further refining the chronological parametersfor the two sites. Renewed excavation at the two sites was alsoundertaken as part of a systematic study by the Sealinks Projectinto the trans-oceanic biological exchange that brought a varietyof Asian plants and animals to the East African coast in thepremodern period (Boivin et al. 2013, 2014; Crowther et al.2014, 2015; Helm et al. 2012). This paper addresses the glasscomponent of the bead assemblages recovered from Fukuchaniand Unguja Ukuu in the 2011 field season.

The rich glass bead assemblage from Zanzibar, the result ofmore intensive recoverymethods than in previous excavationson the island, provides an exceptional opportunity to exploreZanzibar’s connections to the broader Indian Ocean world.Most of the beads recovered are not morphologically distinct;they are small and monochrome and are often types that weremade over the span of several thousand years. Here we at-tempt to explore the beads’ origins through geochemical andtechnological analyses that provide insights into their produc-tion. We perform elemental analyses to trace the origins of thedifferent glasses used to produce the beads. Since glass waswidely traded in antiquity, however, we also examine themethod by which the beads were made. We draw on thesedatasets to begin to reconstruct the trade routes that broughtthe beads to the African shore. We also compare the Zanzibarbead assemblages and the trade connections they inform with

those in southern Africa during the same period to attempt todiscern whether or not the two ends of the eastern Africancoast were involved in overlapping trade circuits.

Material and methods

Sites and samples

Unguja Ukuu is located on a narrow coral-rag peninsula be-tween a resource-rich bay and a small, shallow inlet. The ar-chaeological site extends over some 17 ha and contains settle-ment remains and deep midden deposits rich in animal bone,shell, iron slag, daub, glass fragments and beads, shell beads,bead grinders, and pottery. Fukuchani is situated on a longbeach protected by the island of Tumbatu, which sits directlyopposite the bay. When first recorded by Horton and Clark in1985, the site comprised a series of midden mounds that ran for1 km along the beach, though on returning in 2011, the SealinksProject found that these had been largely destroyed by the con-struction of a local school on the site. Our excavations targetedareas where subsurface deposits appeared intact.

The artifacts analyzed in this study derive from four trench-es at Unguja Ukuu (trenches 10–13) and three at Fukuchani(trenches 10–12). All trenches were between 1× 2 m and2×2 m in area and were 0.7–1.75 m deep when sterile de-posits were reached, though trench 11 at Unguja Ukuureached a maximum depth of 2.92 m. The glass beads werefound throughout the full depositional sequence at both sitesin association with quantities of other imported goods, includ-ing Chinese and Near Eastern ceramics (e.g., Changshapainted stoneware, Yue Green ware, and Dusun stonewarefrom China; and turquoise-glazed and white-glazed waresfrom Iraq and/or Iran) and glass vessel fragments. They alsoco-occur with local Early Tana Tradition/Triangular IncisedWare (ETT/TIW) pottery that dates regionally to the sev-enth–tenth centuries CE (Fleisher and Wynne-Jones 2011).A suite of radiocarbon dates obtained from both sites supportthis chronology (Crowther et al. in preparation). Post-1000 AD ceramics found in the upper levels of trench 13 atUnguja Ukuu, including imported sgraffito ceramics, suggestthat occupation of some of the excavated areas extended intothe eleventh century CE or so. Trench 11 at Fukuchani pri-marily contained a human burial.

All deposits at both sites were sieved with 3 mm or smallermesh resulting in excellent retrieval rates: 863 glass beads andone presumed weight came from Unguja Ukuu and 30 beadscame from Fukuchani (see Online Resource 1 for a full de-scription of the samples and the contexts1 from which they

1 The context of an archaeological object describes where in the excava-tion it was found—including the trench, the section in the trench, and thedepth. This stratigraphic information can help date objects.

880 Archaeol Anthropol Sci (2017) 9:879–901

came). The selection of the beads for chemical analysis wasnot random but rather a range of each type and color waschosen to try to include all possible glass types present.Thus, 68 beads from Unguja Ukuu and 11 beads fromFukuchani were chosen for analysis. A small blue cuboidalglass object from Unguja Ukuu (UU174) measuring5.5×5.5×3 mm and decorated on one side with nine smallindented circles, thought to be a weight for precious metalssuch as silver or gold, was also included in the analysis forcomparison.

Instrumentation

Elemental analysis was performed through laser drilling vialaser ablation on a spot on the glass surface (LA system: NewWave Research UP 213, Fremont, USA) followed by onlineinductively coupled plasma-mass spectrometry (ICP-MS)(quadrupole Agilent 7500ce, Palo Alto, USA). The laser ab-lation device contained a frequency quintupled Nd:YAG laser

(wavelength 213 nm and pulse width 4 ns) with a motorizedstage. Tuning of the LA-ICP-MS system and optimization ofthe sensitivity was carried out using the standard glass NIST612. Sampling of the glass artifacts was conducted using thespot mode with a laser beam diameter of 100 μm and a repe-tition rate of 10 Hz, yielding a penetration rate of ca.1.1 μm s−1. The mass spectrometer was set up in time-resolved analysis mode, measuring one point per isotopicmass and acquiring 53 masses values (Table 1 gives theoperational conditions). In order to establish a blank signalfor all masses, laser ablation started 10 s after the gas blankmeasurement (He/Ar mixture). A more detailed description ofthe LA-ICP-MS system used in this work is given in vanElteren et al. (2009).

Five hundred shots were fired per spot. The raw ICP-MSdata acquired (in counts per second) during the last 100shots were averaged for each of the 53 isotopes.Subsequently, the raw ICP-MS data were subjected toquantification with the so-called sum normalization

Archaeol Anthropol Sci (2017) 9:879–901 881

Fig. 1 Indian Ocean map with places mentioned in the text

calibration method previously described (van Elteren et al.2009) based on summation of the oxides to 100 % and usingSi as an internal standard to correct for ablation differences

between different glass matrices. Four different series ofstandard reference materials were used to determine theconcentration of major, minor, and trace elements: NISTSRM 610 and 612 (National Institute of Standards andTechnology); SGT 2, 3, 4, and 5 (Society of GlassTechnology); CMG B, C, and D (Corning Museum ofGlass); and DLH 6, 7, and 8 (P&H Developments Ltd.).

Glass standards from the Corning Museum of Glass(CMG), which mimic ancient compositions, are the most suit-able for accurate quantification of ancient glasses, especiallyfor elements not present at trace levels, such as in lead glasses.CMGA (no longer available) and B match the composition ofSLS Egyptian, Mesopotamian, Roman, Byzantine, andIslamic glasses. CMGC is a lead-barium glass similar to thosefound in East Asia, and CMGD is a high-Mg, high-Ca potashglass with typical medieval composition (Brill 1999). Sinceancient glass samples often show signs of surface degradation,the laser was used in drilling mode to enable the measurementof the actual elemental bulk composition underlying any de-graded layer. Most of the samples presented leached and cor-roded surfaces with an enrichment of magnesia, potash, lime,or alumina compared to the pristine glass, due to the precipi-tation of mineral phases from the soil onto the glass surface.

Table 1 LA-ICP-MS operating conditions for LA spot analysis drillingprocedure

Laser (NWR UP213)

Beam diameter 100 μm

Fluence 7 J/cm−2

Repetition rate 10 Hz

Penetration rate ca. 1.1 μm s−1

Dwell time 50 s

ICP-MS (Agilent 7500cs)

Number of measuredelements

53 (9Be, 11B, 23Na, 24Mg, 27Al, 29Si, 31P, 39K,43Ca, 45Sc, 47Ti, 51V, 53Cr, 55Mn, 57Fe, 59Co,60Ni, 63Cu, 66Zn, 69Ga, 75As, 82Se, 85Rb,88Sr, 89Y, 90Zr, 93Nb, 95Mo, 107Ag, 111Cd,115In, 118Sn, 121Sb, 137Ba, 139La, 140Ce,141Pr, 146Nd, 147Sm, 153Eu, 157Gd, 159Tb,163Dy, 165Ho, 166Er, 169Tm, 172Yb, 175Lu,197Au, 208Pb, 209Bi, 232Th, 238U)

Total acquisition time(per element)

0.64 s (10 ms per element)

882 Archaeol Anthropol Sci (2017) 9:879–901

Fig. 2 Map of Zanzibar, showingthe location of Fukuchani andUnguja Ukuu

Results

Online Resource 2 presents all the data from the analysis of 80artifacts (but including 84 analyses since several beads aremulticolored and each color was tested) comprising the con-tents of 52 elemental oxides. Online Resource 3 presents re-duced compositions of all samples based on the method ofBrill (1999, pp. 8–11) [in tables and graphs reduced composi-tions are indicated with *]. Online Resource 4 presents meanconcentrations and standard deviations of all oxides by glasstype and color, and Online Resource 1 provides morphologi-cal characteristics of all samples, their site contexts, and glasstype.

Three basic types of soda-silica glass are represented in theassemblages, one in which the flux used in primary glassmanufacture was derived from mineral soda (m-Na-Al) andthe other two where the flux was a halophytic plant ash (v-Na-Ca and v-Na-Al) [the v stands for végétale, French for plant].Magnesia concentrations are useful in separating these glasstypes: mineral soda glasses are normally characterized byMgO levels below 1.5 wt%, while plant ash glasses normallycontain levels greater than 2.5 wt%. In addition, the m-Na-Alglass from these Zanzibar sites also contains higher concen-trations of alumina than the plant ash glasses. Both m-Na-Aland v-Na-Ca glass types were found at Unguja Ukuu andFukuchani but v-Na-Al occurred only at Unguja Ukuu.Figure 3 clearly shows the separation between these glasstypes based on their lime and alumina concentrations.

Discussion of the glass types and the beads produced

Mineral-soda-alumina (m-Na-Al) glass

Fifty-four (68.4 %) of the 79 tested Zanzibar beads are madeof mineral-soda-alumina glass (Online Resources 1 and 4).Dussubieux et al. (2010) have identified five different sub-types of m-Na-Al glass in samples dating from about the fifthcentury BCE to the nineteenth century CE. Two of these (mNa-Al 1 and m-Na-Al 2) are found at sites in Africa, though

the time span during which this glass occurred at sites sampledto date within Africa is more restricted. One of these glasssubtypes, m-Na-Al 1, is present in the Zanzibar beads alongwith a few outliers that do not comfortably fit any knownsubtype. The m-Na-Al 1 glass was formerly known as lowuranium-high barium (or lU-hBa) glass. At first, we thoughtit possible that some of the outliers belonged to the m-Na-Al 2subtype, a high uranium-low barium glass formerly known ashU-lBa (Dussubieux et al. 2008), but that attribution may bequestionable as will be shown.

Dussubieux et al. (2010) differentiated the m-Na-Al 1 and2 subtypes on the basis of the concentrations of four charac-terizing elements: Sr, Zr, Ba, and U. These four elements wereprobably chosen because they were the only elements whichpresented an acceptable separation between the concentrationranges in m-Na-Al 1 and 2 glasses, even though the first threeof these elements partially overlap. An analysis of the wholedataset, however, indicates that alumina may also be useful inseparating the two subtypes, so here we include it and thusemploy five elements—Sr, Zr, Ba, U, and Al—to characterizeand compare the m-Na-Al Zanzibar beads.

In comparing the Zanzibar m-Na-Al glasses to the compo-sitions of Dussubieux et al. (2010) m-Na-Al 1 and 2 subtypes,we used compositional data, kindly provided by Dussubieux.This included beads from South and West India as well asKenya. Our Sri Lankan data came from Dussubieux (2001).The source of the South Indian data included two archaeolog-ical sites that appear to predate the current era. The WestIndian data is from samples ranging in date from the ninth tonineteenth centuries CE obtained at Chaul, a port site south ofpresent-day Mumbai. The African samples are from four sitesin Kenya. One of these has dates that span from the tenth toeighteenth centuries CE, and the others postdate the tenthcentury and cluster between the thirteenth and sixteenth cen-turies. The Sri Lankan data is from Giribawa, a third-centuryBCE to second-century CE glass- and beadmaking center.

When subjected to principal component analysis (PCA),the compositions of these beads (expressed in ppm of theelements and then standardized) formed two clusters represen-tative of the compositional features of most of the samples ofeachm-Na-Al subtype (see blue and pink ovals of Fig. 4). Themajority of the Zanzibari, Sri Lankan, and South Indian beadsform a fairly compact group that suggests they are closelyrelated, but the numerous outliers from this core suggest thatthere may be other subtypes that have not yet been identified.The overlapping between the two subtypes is a consequenceof samples that lie between the concentration ranges as report-ed by Dussubieux et al. (2010).

The PCA extracted five principal components, of whichonly the first two (PC1 and PC2) were considered significanton the basis of the scree plot. These explain about 70 % of thevariance. Most of the Zanzibar beads (red points of Fig. 4)form a compact cluster within the area of Dussubieux et al.’s

Archaeol Anthropol Sci (2017) 9:879–901 883

Fig. 3 Plot of lime vs alumina for all analyzed Zanzibar beads (reducedcompositions calculated using the method of Panighello et al. (2012)

m-Na-Al 1 subtype, while the samples reported in Table 2appear external to or borderline to this group. In particular,samples UU003, UU097, UU144, and UU163 fall in the areaof overlap between the two subtypes. The analysis of variableloading (Fig. 4 inset) shows that along PC1 uranium and theother four elements as a group are inversely correlated: theformer (with highly negative loading) essentially favors them-Na-Al 2 cluster, while the latter load on the m-Na-Al 1 side.The variable loadings of the second PCA component showthat the positions of the single samples reported in Table 2(except UU003, 097, 144, and 163) are due mainly to theloading of the Al, Sr, and U variables.

The blue and pink ovals of Fig. 4 statistically circumscribeabout 70 % of the m-Na-Al 1 and m-Na-Al 2 samples on thebasis of the Dussubieux data. Therefore, the borderline sam-ples (UU013, 057, 059, 212, 218, and 228) and those justoutside the border (UU204, 209, and 213) may be part ofthe m-Na-Al 1 subtype. As Table 2 illustrates, none of thebeads in the area of overlap between the two subtypes fitunequivocally into the m-Na-Al 2 category. Finally, althoughthe PCA analysis represents a statistical synthesis of the con-centration distributions of five elements, a detailed analysis ofthe data for the samples individually reported in Table 2 (andillustrated in Fig. 4) indicates the contradictory or uncertainassignments of these samples when the data for each elementare considered separately. This underscores the limits of theDussubieux et al. (2010) criteria when attempting to deter-mine to which subtype some individual samples belong. In

concurrence with this observation, new research byDussubieux on other Zanzibar beads has confirmed the diffi-culty of assigning beads of m-Na-Al glass from the EastAfrican coastal region to only subtypes 1 and 2 (2015, per-sonal communication).

Unfortunately, due to their similarity, beads of the two m-Na-Al subtypes cannot be separated based on morphology. Itis interesting, however, that the majority of the chemicallyunusual beads that do not fit comfortably in either categorycome from disturbed contexts. As Table 3 shows, 8 of the 13beads were found in trench 10 contexts 005, 006, and 009—all of these are from the fill of a pit dug into the trench.

m-Na-Al 1

Glass of the m-Na-Al 1 subtype is known from Sri Lankan,South Indian, and Southeast Asian sites dating between aboutthe fifth century BCE and the tenth century CE (Dussubieux etal. 2009, p. 159, 2010; Dussubieux 2001; Dussubieux andGratuze 2013). Archaeological and associated chemical evi-dence indicates that one manufacturing center for this glasswas located at Giribawa in Sri Lanka (Bopearachchi 1999,2002; Dussubieux 2001), but that site is dated between thethird century BCE and the second century CE and is thustoo early to have been the source of the Zanzibar glass. Noother sites manufacturing this glass type have been located,but there likely were other such glassmaking centers in SriLanka; for example, Francis (2013) concluded that Mantai

884 Archaeol Anthropol Sci (2017) 9:879–901

Fig. 4 PCA of m-Na-Al samples based on U, Ba, Zr, Sr, and Al levels.The Zanzibar beads are compared to others from Sri Lanka (Dussubieux2001) and South India, West India, and Kenya (data courtesy ofDussubieux; see also Dussubieux et al. 2008, 2010). The black ovaldelineates samples from Sri Lanka, while the gray dashed-dottedrectangle outlines the South Indian ones. The red dashed line includes

all of the Zanzibar samples. The blue circle delineates the Dussubieuxet al. (2010) m-Na-Al 2 group parameters, and the pink oval the param-eters of their M-Na-Al 1 group. Individual Zanzibar beads that do not fallcomfortably within the parameters of either m-Na-Al 1 or 2 are identifiedby ID number and their relevant element concentrations and affinities arelisted in Table 2

was both a glassmaking and beadmaking center. However, asDussubieux and Gratuze (2013, p. 404) note, South India may

have produced m-Na-Al 1 glass as well. The orange beadsmade of this glass (a rather bright pumpkin orange [Munsell5YR 5/10]) are potentially significant markers in that this isthe first time they have been recorded in eastern Africa, al-though they do occur in the late occupation phase (fourth tosixth centuries CE) at Berenike, a Red Sea port in Egypt(Francis 2002, p. 228, note 21). They were common atMantai where Francis (2002) proposed they were made. It isnoteworthy that this is the first time m-Na-Al 1 glass has beenrecorded in eastern or southern Africa, apart from two mor-phologically unusual beads: one from Ungwana on theKenyan coast (Dussubieux et al. 2008, p. 814) and the otherfrom Mahilaka in northwest Madagascar (Robertshaw et al.2006).

m-Na-Al 2

Glass of the m-Na-Al 2 subtype is widely distributed in east-ern and southern African sites, as well as at sites in

Table 2 Average concentrationsand standard deviations in partsper million or weight percentageof elements/oxides important forseparating m-Na-Al 1 from m-Na-Al 2 glass (data fromDussubieux et al. 2010), followedby data for questionable samplesfrom Unguja Ukuu withsuggestions of possible subtypeattribution based on each element/oxide

Sr (ppm) Zr (ppm) Ba (ppm) U (ppm) Al2O3 (wt%)

m-Na-Al 1 Dussubieux et al. (2010) 373 ± 145 561 ± 420 931 ± 432 11 ± 10 9.8 ± 2.1

m-Na-Al 1 Our data 345 ± 58 538 ± 133 826 ± 210 15± 18 9.3 ± 1.4

m-Na-Al 2 Dussubieux et al. (2010) 213 ± 70 168 ± 91 357 ± 114 105 ± 66 7.5 ± 1.6

m-Na-Al ?

UU003 235 ± 16 540 ± 42 872 ± 46 121 ± 9 10.8 ± 0.2

Related to subtype: 2 1 1 2 1

UU013 273 ± 4.3 922 ± 16 985 ± 15 7 ± 2 6.6 ± 0.1

Related to subtype: 2(?) ? 1 2 2

UU097 215 ± 6 293 ± 8 438 ± 12 24± 1 8.5 ± 0.1

Related to subtype: 2 1 2 1 1 or 2

UU144 233 ± 4 315 ± 7 483 ± 15 29± 1 8.4 ± 0.1

Related to subtype: 2 1 2 1 1 or 2

UU163 232 ± 3 321 ± 6 480 ± 10 27± 1 8.4 ± 0.1

Related to subtype: 2 1 or 2 2 1 1 or 2

UU213 567 ± 6 1400± 45 1218 ± 13 20± 1 12.1 ± 0.1

Related to subtype: ? ? 1 1 1

UU218 589 ± 20 1021± 58 1260 ± 25 19± 1 12.1 ± 0.1

Related to subtype: ? ? 1 1 1

UU057 589 ± 6 674 ± 7 1263 ± 35 18± 1 13.1 ± 0.3

Related to subtype: ? 1 1 1 ?

UU212 460 ± 4 277 ± 3 587 ± 6 94± 2 10.8 ± 0.1

Related to subtype: 1 1 or 2 1 2 1

UU228 459 ± 14 277 ± 10 572 ± 20 95± 3 10.8 ± 0.1

Related to subtype: 1 1 or 2 1 2 1

UU059 626 ± 19 331 ± 17 633 ± 37 36± 2 11.8 ± 0.2

Related to subtype: ? 1 1 1(?) 1

UU204 720 ± 20 204 ± 5 806 ± 15 59± 2 12.2 ± 0.1

Related to subtype: ? 1 or 2 1 2 1

UU209 368 ± 25 133 ± 8 344 ± 17 88± 3 11.3 ± 0.1

Related to subtype: 1 2 2 2 1

Table 3 Trench andcontext of beads whosesubtype designation isuncertain

Bead ID Trench Context

UU209 10 009

UU097 10 005

UU163 10 006

UU144 10 006

UU213 10 009

UU218 10 009

UU204 10 009

UU212 10 009

UU228 13 006

UU057 11 017

UU059 11 017

UU013 11 005

UU003 11 002

Archaeol Anthropol Sci (2017) 9:879–901 885

Madagascar and India and beyond, dating from about the mid-tenth to seventeenth centuries CE (Dussubieux et al. 2008;2010; Robertshaw et al. 2003, 2006, 2010b). In East Africa,some reports indicate that this glass type appears in sites thathave date ranges that include the ninth century (Dussubieuxand Gratuze 2013, p. 404) but no details of the contexts inwhich the beads were found are provided. In addition, theearlier East African dating evidence is probably less securethan that for southern Africa. All available evidence indicatesthat m-Na-Al 2 glass was probably manufactured at a numberof locations in India widely distributed from the Uttar Pradeshregion southwards from a mineral soda known, at least inrecent times, as reh (Brill 2001a, b; Kanungo 2004; Sodeand Kock 2001). It appears unlikely that any of the Zanzibarbeads were made from this glass type. As Table 2 and Fig. 4demonstrate, none of the beads fit unequivocally into thattype’s parameters. In addition, the AMS dates for the trenchesin which potential examples were recovered predate the tenthcentury, whereas this glass type has not been recorded earlierthan the tenth century.

The most important conclusion that can be drawn from theanalysis of the m-Na-Al beads found in Zanzibar is the pro-nounced contrast between the prevalence of subtype 1 andprobable absence of subtype 2 of this glass in the assemblageand the fact that virtually all the m-Na-Al glass from southernAfrica belongs to subtype 2 (Robertshaw et al. 2010b). Thesedifferences can be explained by temporal parameters: subtype1 glass, as has been mentioned, was produced up to but notbeyond the tenth century CE (Dussubieux et al. 2009, p. 159,2010; Dussubieux 2001; Dussubieux and Gratuze 2013),while in southern Africa, no m-Na-Al glass beads have beenrecorded that predate the tenth century (Robertshaw et al.2010b; Wood 2011).

Plant ash soda glasses

Two distinct types of glass were identified among the 26 sam-ples fluxed with plant ash. The first type (v-Na-Ca) comprises23 samples (including the presumed glass weight) that can betentatively divided into two subtypes and two outliers, whilethe second (v-Na-Al), which contains elevated levels of alu-mina, is represented by only three beads; these will bediscussed separately.

v-Na-Ca

All of the 23 v-Na-Ca samples have relatively small quantitiesof alumina. Twelve of the beads, along with the blue and whiteglasses of the eye bead (UU225), and the glass weight, form asubtype, here designated A, characterized by medium concen-trations of lime, while subtype B, characterized by higherconcentrations of lime, is represented by seven beads plusthe black base glass of the eye bead (UU225). Two outliers,

UU091 and UU173 with their higher soda concentrations andmostly lower concentrations of other major oxides (as well astrace element concentration differences), do not fit any of theidentified groups; these are discussed in the section BTheoutliers^ below.

The distinction between the proposed subtypes is tentative.Lime and phosphorus pentoxide contents show quite goodseparation between the subtypes (Fig. 5), but samples cannotbe easily assigned to a subtype based on the quantities ofnumerous other elements, though some elements, such aschromium and boron, appear more promising as discriminants(e.g., Fig. 6).

v-Na-Ca subtype A Subtype A includes 12 beads, the blueand white glasses of the eye bead (UU225) and the glassweight but produced 17 analyses because three beads are mul-ticolored and each color was tested; the multicolored beads arediscussed more fully below. Colors that are found in subtypeA beads include cobalt blue, blue-green colored with copper,white colored with tin and one colorless sample (UU114) witha notable, if predictable, absence of any coloring agent. Fourof the seven cobalt blue beads, including the two stripedbeads, have been reheated on a flat surface (see BBead typesin the overall 2011 Zanzibar assemblages^ for a descriptionand Online Resource 1 for details). Among the blue-greenbeads, UU090 was pinched from a large tube with a largeperforation and then reheated on a flat surface. FK017 is sim-ilar and may have been produced in the same manner. Theremaining two, UU112 and UU235, are smaller but still havelarge perforations. Their walls are very thin so the ends havedeteriorated making it impossible to determine whether theywere reheated at all. The colorless bead (UU114) is uniformly

886 Archaeol Anthropol Sci (2017) 9:879–901

Fig. 5 Plot of lime vs phosphorus pentoxide for the v-Na-Ca A and Bsamples. Measurements in parts per million

rounded on both ends. The remaining v-Na-Ca subtype Asamples are the white and blue Beyes^ within the eye beadwhich is the only wound bead in the Zanzibar assemblages; itis described in the following section.

In the Zanzibar assemblages, cobalt blue glass is foundonly in beads (and the weight) made of v-Na-Ca glass andspecifically only within subtype A; those with cobalt concen-trations of over 1000 ppm (n= 9) will be discussed here.Cobalt in glass is often associated with various elements in-corporated in the raw cobalt-bearing materials. Gratuze et al.(1992) note that cobalt can come from cobaltite (CoAsS),skutterudite (Co,Ni) As3-x, or trieuite (2Co2O·CuO·6H2O),or it can be associated with Fe, Ni, and As or Pb, Zn, In, etc.One of the three main cobalt mineral groups identified byGratuze et al. is cobalt-zinc, in which these two elements arecorrelated with lead and indium. The cobalt in the Zanzibarbeads is correlated with zinc and copper but not with otherelements; they do include variable amounts of lead (0.1 to5 wt%) and iron (0.9 to 2.1 wt%) but these are not correlatedwith the cobalt, and our indium levels are unfortunately notreliable due to the interferences between 118Sn and 115Inisotopes in the ICP-MS system. Figure 7 shows that there isa strong Co-Zn linear correlation for all of our blue samplesexcept UU225 (the eye bead), which has a Co/Zn ratio of 0.8compared to the other samples with a ratio of 3. This differentglass is the blue Bpupil^ of the eye bead, and given that themorphology and construction technology of this bead is sodifferent from all the others, it might not be surprising that aglass colored with cobalt from a different source was used inits manufacture. Unfortunately, there is a dearth of informationon the chemistry of cobalt sources in the Middle East duringthe early Islamic period.

Unguja Ukuu produced three multicolored beads, all fromtrench 10 below the rubbish pit mentioned earlier (section Bm-Na-Al 2^). Two of these beads are made of v-Na-Ca subtypeA glass, while the third (UU225) consists of a base glassassigned here to v-Na-Ca subtype B glass and eyes of blueand white glass that appear to be of subtype A. The fact thatboth v-Na-Ca subtypes appear to have been used in the man-ufacture of a single bead highlights the tentative nature of theidentification of two subtypes; however, it has been shownthat glass workshops in the early Islamic period sometimesused raw glass from different primary sources (Freestone etal. 2002, p. 270; Robertshaw et al. 2010b: Table 7), so it iswell within the bounds of possibility that two different(sub)types of glass could have been used in the manufactureof a single multicolored bead.

Studies of these multicolored beads and others like themfrom sites in Sri Lanka, Thailand, and Scandinavia have en-abled their manufacture to be placed in a time frame roughlybetween the late eighth and mid-ninth centuries (Wood, un-published data). Two of these (UU231 and UU232) are cobaltblue with white stripes that run parallel to the perforation. Thethird bead (UU225) is known as a stratified eye bead (seeFig. 17 for images). It is wound (all others in the entireZanzibar assemblage are drawn) with a black body and eyesmade of a circle of white glass topped by a smaller one ofcobalt blue. Eye beads are fairly common with a very longhistory, but the type discussed here is specific enough to havebeen named the BTakua Pa eye bead^ after an eighth- toeleventh-century archaeological site (sometimes calledTakua Pa but actually named Thung Tuk) on the Andamancoast of peninsular Thailand (Chaisuwan 2011) where a largenumber were found and where Francis (2002, p. 97) proposedthey may have been made. We compared the chemistry ofUU225 with three other BTakua Pa^ eye beads from ThungTuk (data courtesy of Jim Lankton), an eye bead from al-Basrain Morocco (PR1019FM; Robertshaw et al. 2010a) and two

Archaeol Anthropol Sci (2017) 9:879–901 887

Fig. 6 Plot of chromium vs boron for the v-Na-Ca A and B samples.Measurements in parts per million

Fig. 7 Correlation between zinc and cobalt in the blue Zanzibar beads

eye beads from Chibuene in Mozambique (CHB0045 andCHB0047; Wood et al. 2012). All of these sites have eighth-to ninth-century components.

The al-Basra and Chibuene eye beads are morphologicallyunlike the BTakua Pa^ eye beads but all are made of v-Na-Caglasses, as is UU225. As Fig. 8 demonstrates, UU225 is verysimilar in terms of its major oxide concentrations and ratios tothe Thung Tuk beads, being characterized by a high combinedconcentration of magnesia and potash and a low alumina tolime ratio. By contrast, the Chibuene eye beads differ in theiralumina to lime ratios, while the al-Basra bead has less mag-nesia and potash than the Zanzibar and Thung Tuk specimens.The distinction between the Zanzibar eye bead and those fromThung Tuk on the one hand and the al-Basra eye bead on theother seems to be borne out by examination of the apparentcoloring agents in the white glasses of these beads as well as ofthe blue and white striped beads from Unguja Ukuu andTumbe, discussed next. The al-Basra white glasses, but notthe others, are notable for their relatively high concentrationsof lead and arsenic, as well as tin (Table 4), suggesting that thewhite color in the al-Basra glass might derive from a mixtureof cassiterite and lead arsenate, whereas the others, includingthe Chibuene eye beads, seem to have been colored primarilywith cassiterite. It is perhaps worth noting that the lead/arsenicratios of the al-Basra white glasses are very different fromthose of the other white glasses discussed here. Moreover,with the exception of al-Basra, the quantities of arsenic, tin,and lead in the white glasses from the other sites are notcorrelated.

The two striped beads from Unguja Ukuu (UU231 andUU232) are morphologically very similar to one (PR788)

from the site of Tumbe on the north end of Pemba Islandin Tanzania (Fleisher and LaViolette 2013). Similar beadsare also present at Thung Tuk. All these beads are alsochemically very similar, with the exception of the blue glassof the Tumbe bead (Table 5 and Fig. 9), but this particularglass is clearly corroded, as is evident from both its highsilica and low soda levels. In all the beads, the blue colorcan be ascribed to additive levels of both copper and cobalt,while the white glass exhibits high levels of tin, corre-sponding to tin oxide used as a white opacifier. The ratiosof lead and tin to arsenic are strikingly similar in the whiteglasses of UU232 and the Tumbe bead (Table 4). Althoughall of these sites have eighth- to ninth-century components,this type of striped bead may also be present outside thistime frame.

v-Na-Ca subtype B Of the eight subtype B beads, five are anopaque brick-red glass that is filled with bubbles, giving theends a sponge-like appearance (see Fig. 17 FK019a and b): thethree from Fukuchani (FK019a, b, and c) are medium sized(5.5 to 6 mm diameter), while the two from Unguja Ukuu(UU007 and UU115) are large (9.6 diameter) and appear tohave been pinched from a glass tube rather than cut from tubeslike the smaller beads. All were reheated on a flat surface toslightly round the cut ends which is an unusual treatment (seesection BBead types in the overall 2011 Zanzibarassemblages^). Two subtype B beads are translucent blue-green and large (the more complete one measures 11 mm indiameter), and they were made by a process known assegmenting. Both the brick-red and blue-green subtype Bbeads were colored with high concentrations of copper. The

888 Archaeol Anthropol Sci (2017) 9:879–901

Fig. 8 Reduced alumina (Al2O3)and magnesia (MgO)compositions (wt%) of UU225compared with those of other eyebeads. The three data points forUU225 represent the threedifferent colored glasses that wereanalyzed on this bead. The otherbeads in the figure also havemultiple data points, each basedon the analysis of a differentlycolored glass

final bead of this subtype comprises the black glass that formsthe base of the eye bead (UU225) discussed above.

The origins of the increased concentrations of CaO in thesubtype B glass are not entirely clear. Henderson (2013, p.284) reports that most lime in plant ash glass comes fromthe ash itself. However, the lime content of some halophyticplants was insufficient to make Islamic glass, so a third pri-mary raw material, such as feldspar which is rich in bothcalcium and alumina, could have been added along with thesand which, especially with desert and stream sands, is oftenlow in calcium (Duckworth et al. 2015). Unfortunately, ourdata cannot differentiate between potential calcium sources.The mean ratios between the lime, alumina, and phosphorusoxide contents in subtypes A and B are similar, but otheroxides that characterize the ashes, such as sodium, potassium,and magnesium, are rather different. In the end, these differ-ences cannot account for the differences in lime levels be-tween the two subtypes, thus suggesting at the very least theuse of ashes from different plants.

Comparison of v-Na-Ca subtypes to related datasetsAfrican comparisons: There are two datasets comprisingchemical compositions of beads of plant ash glass fromthe late first millennium CE that are appropriate for com-parison with the Zanzibar v-Na-Ca subtypes. The first com-prises the plant ash glass beads from the port of Chibuene inMozambique where three subtypes of plant ash glass wereidentified (Wood et al. 2012). The first of these, ChibueneBv-Na 1^ glass, is represented by beads, vessel shards,blobs, and cullet, with the beads identified as belonging tothe Zhizo series (Wood 2011). Chibuene Bv-Na 2^ glass,with relatively high concentrations of chromium, nickel,and zirconium, consists of green vessel shards, cullet, andblobs but no beads, while Chibuene Bv-Na 3^ glass is rep-resented only by beads which were assigned to a new mor-phological type, the Chibuene series (Wood et al. 2012).The second comparative African dataset used here com-prises the results of chemical analysis of 16 beads of theZhizo series from various southern African sites

Table 4 Concentrations (in ppmor weight percentage as shown)and ratios of the coloring agentsin the white glasses of the eye andstriped beads

ANID Site As2O3 SnO2 (%) PbO (%) PbO/As2O3 SnO2/As2O3

PR1019FM al-Basra 0.14 % 16.89 10.28 72 118

PR1019FM-2 al-Basra 0.12 % 26.85 11.78 95 217

UU225 Unguja Ukuu 88.58 3.07 2.27 297 346

KKK TT 041 Thung Tuk 10.55 7.44 1.22 1161 7047

KKK TT 040 Thung Tuk 9.93 5.54 0.95 957 5581

KKK TT 058 Thung Tuk 12.90 7.85 0.51 398 6086

CHB0045 Chibuene 0.00 2.43 2.92 – –

CHB0047 Chibuene 76.46 3.09 1.99 197 405

UU231 Unguja Ukuu 36.71 2.63 0.15 41 717

UU232 Unguja Ukuu 49.30 6.15 1.40 285 1247

KKK TT 044 Thung Tuk 12.41 4.33 0.91 732 3487

KKK TT 036 Thung Tuk 119.94 7.40 2.31 193 617

PR788 Tumbe 34.08 3.56 0.88 259 1043

Table 5 Reduced compositions (weight percentage) of the striped beads. Reduced compositions are calculated following the method described byBrill (1999, pp. 8–11)

Bead ID Site Color Na2O* MgO* Al2O3* SiO2* K2O* CaO* Fe2O3*

KKK TT 044 Thung Tuk Blue 15.2 4.1 1.7 68.9 3.2 5.4 1.5

KKK TT 044 Thung Tuk White 15.4 4.7 2.2 67.0 3.1 6.4 1.1

KKK TT 036 Thung Tuk Blue 15.0 4.3 2.0 68.4 3.2 5.9 1.2

KKK TT 036 Thung Tuk White 15.4 5.2 2.3 66.4 2.9 7.1 0.8

UU231 Unguja Ukuu Blue 15.2 4.7 2.0 67.2 3.5 5.8 1.6

UU231 Unguja Ukuu White 14.3 4.4 2.1 69.4 3.0 6.0 0.8

UU232 Unguja Ukuu Blue 14.6 4.5 2.2 66.5 3.3 7.4 1.5

UU232 Unguja Ukuu White 15.4 4.8 2.3 65.7 3.6 7.3 1.0

PR788 Tumbe Blue 6.9 3.2 6.4 75.3 1.5 4.6 2.0

PR788 Tumbe White 13.0 3.9 3.5 72.2 2.6 4.3 0.5

Archaeol Anthropol Sci (2017) 9:879–901 889

(Robertshaw et al. 2010b). These Zhizo series beads arechemically very similar to those of the Chibuene v-Na 1series. It may be noted that glass beads with chemistriesvery similar to the Chibuene v-Na 1 and Zhizo series beadshave also been recovered from several West African sites,including Igbo-Ukwu in Nigeria (Robertshaw, unpublisheddata), Gao in Mali (Cissé 2011; Cissé et al. 2013), and Kissiin Burkina Faso (Robertshaw et al. 2009).

Table 6 presents the mean compositions of the major ox-ides and selected trace elements for the Zanzibar plant ashsubtypes and the African comparative dataset. Perusal of thistable and Fig. 10 confirms the distinctively different compo-sition of the Chibuene v-Na 3 glass (Chibuene bead series), asreflected, for example, in the quantities of lime, potash, mag-nesia, chromium, rubidium, barium, and uranium. The secondmost distinctive glass is the Zanzibar subtype B glass with itshigh lime levels, as well as somewhat elevated levels of sodaand chromium, though, as noted above, lime aside, it showsconsiderable compositional similarities to Zanzibar subtypeA, as well as to the Chibuene v-Na 1 and Zhizo glasses.Chibuene v-Na 2 glass is also not entirely dissimilar to theother types, at least when the quantities of individual oxidesare compared, though it has notably higher amounts of chro-mium, nickel, and zirconium and less rubidium. Chromiumand nickel levels in the Chibuene v-Na 2 glass are highlycorrelated (Wood et al. 2012, p. 64), as they are as well inZanzibar subtype B glass (r= 0.855, p=0.007) but not inZanzibar subtype A (r=0.324, p=0.204), though the overallquantities of these elements are generally higher in theChibuene v-Na 2 glass.

Middle Eastern comparisons: Plant ash glasses found ineastern and southern Africa that are dated to the late firstmillennium AD and contain less than about 4 % alumina arewidely considered to have been manufactured in the MiddleEast (Robertshaw et al. 2010b; Wood et al. 2012), though thebeads found on Zanzibar that were made of Middle Easternglass may well have been manufactured outside the MiddleEast (see discussion in section BEast Africa^). The questionhere is whether we can source either of the Zanzibar plant ashglass subtypes to any particular region within the Middle East.Henderson (2013, pp. 290–300) has investigated the complexvariation in glass compositions across the Middle East, notingthe inevitable complexities that result where raw materials, aswell as raw and scrap glass were widely traded, as was thecase here, but nevertheless noting some potential avenues ofinquiry.

Al-Raqqa in Syria is the only primary plant ash glassmanufacturing site of the late first millennium AD that hasbeen thoroughly investigated and for which considerablecompositional data are available (Henderson et al. 2004).Henderson and colleagues identified three plant ash glasstypes (types 1, 2, and 4) at al-Raqqa, with type 4 glassesvarying considerably in composition because they were per-haps the products of experiments in the production process.Figures 11 and 12 compare some of the major oxide compo-sitions of the Zanzibar plant ash subtypes with the al-Raqqatypes. The Zanzibar plant ash glasses fall within the broadcompositional range of al-Raqqa type 4 glasses based on theiralumina and magnesia compositions (Fig. 11), but their potashand lime concentrations tell a different story, with the Zanzibar

890 Archaeol Anthropol Sci (2017) 9:879–901

Fig. 9 Reduced compositions(wt%) of the striped beads fromUnguja Ukuu (Zanzibar), Tumbe(Tanzania), and Thung Tuk(Thailand)

subtype A glasses aligning with al-Raqqa type 1, rather thantype 4, because of their high lime concentrations, though sub-type A still broadly matches al-Raqqa type 4 (Fig. 12).Furthermore, a plot of lime and alumina concentrations inraw furnace glass from al-Raqqa (Henderson 2013, p. 285)shows that this glass is characterized by >6 % lime and<2 % alumina, which makes it a poor match for theZanzibar plant ash glasses.

Looking further afield in the Middle East than al-Raqqa,Henderson (2013, pp. 294–295) has drawn attention to possi-ble broader geographical patterns in the major oxide compo-sitions of Islamic glass across theMiddle East. Given that highlime concentrations are what defines our Zanzibar subtype Bglass, it is noteworthy that comparable high lime glasses in-clude the raw glass from the eleventh-century Serçe Limanishipwreck (Brill 2009), eighth–eleventh-century glass from

Ramla in Israel (Freestone et al. 2000), glass from furnaceson the island of Tyre (Freestone et al. 2002), and glass fromFustat in Egypt (Brill 1999). Henderson (2013, p. 294) ar-gues that the high lime concentrations in these glasses de-rive from shell fragments in beach sands, with theLevantine coast being the most likely source. This encour-ages us to suggest that the Zanzibar plant ash subtype Bglasses most likely derive from the same Levantine region,though it is important to note that the Zanzibar subtype Bglasses generally possess more alumina than do theirLevantine counterparts, suggesting the use of sand ratherthan quartz as the silica source. We should also rememberthat the Levantine samples derive from raw glass and vesselglass rather than colored beads.

It is tempting to look eastwards, to Iraq and Iran, for anorigin for the Zanzibar subtype A glass. Figures 13 and 14

Table 6 Major oxide and selected element compositions (mean and standard deviations) of late first millennium eastern and southern African plant ashglass types

Glass type

Zanzibar A Zanzibar B Chibuene v-Na 1 Chibuene v-Na 2 Chibuene v-Na 3 Zhizo seriesn 17 8 52 13 11 15

SiO2 (%) Mean 65.12 59.70 66.31 67.39 66.37 67.42

SD 2.83 2.50 3.88 1.33 2.01 3.64

Na2O (%) Mean 14.53 15.79 13.87 14.13 15.10 12.66

SD 2.13 0.38 1.01 0.81 0.76 1.81

CaO (%) Mean 6.17 9.46 5.62 5.47 3.68 5.45

SD 0.67 0.64 1.10 0.95 0.33 1.38

K2O (%) Mean 3.03 2.65 3.20 2.97 4.87 3.05

SD 0.35 0.26 0.54 0.36 0.29 0.62

MgO (%) Mean 4.03 3.86 3.75 4.53 2.34 4.20

SD 0.83 0.32 0.49 0.64 0.17 1.37

Al2O3 (%) Mean 2.00 3.10 3.09 2.02 3.86 3.13

SD 0.32 0.69 0.72 0.59 0.38 0.58

Cr (ppm) Mean 53 92 47 191 20 42

SD 9 23 10 56 3 15

Ni (ppm) Mean 30 49 41 115 18 39

SD 9 12 41 44 5 33

Rb (ppm) Mean 15 16 20 10 52 26

SD 2 2 3 2 7 14

Zr (ppm) Mean 135 83 74 288 78 88

SD 45 28 44 130 12 40

Ba (ppm) Mean 148 209 244 170 705 275

SD 74 109 68 57 160 99

La (ppm) Mean 8 8 6 4 16 7

SD 2 1 1 1 2 1

Nd (ppm) Mean 7 7 5 3 13 6

SD 1 1 1 1 2 1

U (ppm) Mean 0.7 0.9 0.8 1.0 3.0 0.7

SD 0.1 0.1 0.2 0.3 1.0 0.3

Archaeol Anthropol Sci (2017) 9:879–901 891

compare the Zanzibar plant ash subtypes with a variety ofIranian and Iraqi glasses of similar age or, in the case of theSasanian glass, earlier. While the high lime and aluminaconcentrations of most of the Zanzibar subtype B glassesshow that they are not a good match with these easternMiddle Eastern glasses, those of Zanzibar subtype A are a

better fit, particularly perhaps with the Sasanian type 1glasses. Freestone (2006) noted that glasses found to the eastof the Euphrates have higher potash and magnesia concentra-tions than those to the west (see also Rehren and Freestone2015: Fig. 5), providing further support for an eastern originfor the Zanzibar subtype A glasses.

892 Archaeol Anthropol Sci (2017) 9:879–901

Fig. 10 Biplot of weight %alumina vs lime in the Zanzibarand other eastern and southernAfrican plant ash glasses of thelate first millennium AD

Fig. 11 Biplot of weight %alumina vs magnesia in theZanzibar and al-Raqqa plant ashglasses (data from Hendersonet al. 2004)

In summary, we suggest that the Zanzibar plant ash sub-type A glasses may have been made east of the Euphrateswhile those of subtype B more likely originate in theLevantine region. These ascriptions are tentative. More re-liable sourcing of the glasses within the Middle East willrequire much more detailed investigation of both trace ele-ment and isotopic data.

The outliers The two beads that are outliers (UU091 andUU173), which are made from v-Na-Ca glass but with ratherdifferent chemistries, remain to be discussed. UU173 is blue-green, colored with copper, while UU091 is blue but not darkcobalt blue like the other blues—it contains both copper andcobalt with a Cu/Co ratio of about 3 which results in a bluetending toward green tones rather than purple ones.

Archaeol Anthropol Sci (2017) 9:879–901 893

Fig. 12 Biplot of weight %potash vs lime in the Zanzibar andal-Raqqa plant ash glasses (datafrom Henderson et al. 2004)

Fig. 13 Biplot of weight %alumina vs magnesia in theZanzibar and eastern MiddleEastern plant ash glasses (datafrom Brill 1995, 1999; Mirti et al.2008, 2009)

UU091 is, in terms of its major oxides, broadly comparableto Zanzibar subtype B but includes more soda and less of theother oxides. It also exhibits some differences in trace elementconcentrations from the Zanzibar subtype B glasses, for ex-ample, the vanadium, zinc, silver, tin, and barium concentra-tions. UU091 also possesses a very distinct rare earth element(REE) profile (Fig. 15), which distinguishes it from otherMiddle Eastern glasses. This REE pattern, with a high enrich-ment of heavy REEs, is typical of cobalt blue glasses derivedfrom cobalt alum sources, as found, for example, in LateBronze Age Egypt (Shortland et al. 2007).

UU173 has an unusual major oxide chemistry with lowlevels of both alumina and lime, which align the bead withthe glass of the Chibuene series (Wood et al. 2012). However,

the trace element concentrations in UU173 show both similar-ities to and marked differences from those of that series, with,for example, similar concentrations of lime and zirconium butdifferent quantities of alumina, chromium, and rubidium(Fig. 15; Table 7), presumably indicating closely related butnot identical glasses. Wood et al. (2012, p. 66) remarked thatthe chemistry of the Chibuene series beads is Bstrikingly dif-ferent from any of the assemblages in the Middle East,^ aconclusion that we can only echo for UU173.

v-Na-Al

The three plant ash glass beads with high alumina concentra-tions (v-Na-Al) were all found at Unguja Ukuu: UU001 is

894 Archaeol Anthropol Sci (2017) 9:879–901

Fig. 14 Biplot of weight %potash vs lime in the Zanzibar andeastern Middle Eastern plant ashglasses (data from Brill 1995,1999; Mirti et al. 2008, 2009).Note that one Sasanian type 1sample with 8.9 % K2O wasomitted from the plot

Fig. 15 Average REE patterns ofthe two outlier Zanzibar plant ashglass beads, compared with thoseof the Chibuene series (Woodet al. 2012), Zanzibar v-Na-Aland Zanzibar v-Na-Ca subtype Anormalized to REE in thecontinental Earth’s crust(Wedepohl 1995)

yellow, UU002 blue-green, andUU227 green. The differencesbetween this v-Na-Al glass and the other plant ash glassesfrom Zanzibar are demonstrated in their REE profiles(Fig. 16). Subtypes A and B derive from very similar bedrockgeology, whereas v-Na-Al is markedly different, notably withits large positive Eu anomaly, its lack of a negative Ce anom-aly, and its overall higher concentrations of REE.

Two of these v-Na-Al beads came from trench 11 in con-texts 001 and 002, which are made up of overburden mixedwith modern intrusions, while the third, UU227, is fromtrench 13 context 006, which is over half way down the trenchdepth. These beads have a chemistry that is somewhat relatedto that of theMapungubweOblate and Zimbabwe series beadswhich are found at thirteenth- to fifteenth-century sites insouthern Africa and Madagascar (Robertshaw et al. 2006,2010b), as well as to that of vessel glass from Mtwapa onthe southern Kenyan coast (Dussubieux and Kusimba 2012),but they all vary enough to make direct correlations difficult atthis point (Dussubieux, 2015, personal communication).

Thus, based on our present knowledge of this glass type, theseZanzibar beads may represent an as yet unclassified variationof this glass type.

Bead types in the overall 2011 Zanzibar assemblages

It is relatively easy to separate the m-Na-Al and v-Na-Cabeads in these assemblages based onmorphology. First, cobaltblue does not occur in m-Na-Al 1 glass (Dussubieux et al.2008, p. 815, 2010, p. 1650), so all beads of this colorbelong to the v-Na-Ca group. In addition, this type ofglass is often filled with rows of small bubbles alignedparallel to the perforation. When these bubble rows arenear the surface of the bead, they lead to the appearanceof striations (Fig. 17: UU115 and UU007), and when seenin cross section, the bead ends can appear porous andsponge-like (Fig. 17: FK019a and b). Drawn beads aregenerally made by cutting bead-length segments from aglass tube. This results in beads with sharp, even jagged,ends so they are normally reheated to round or smooth theends. The most common way this was done during thisperiod—especially with South Asian beads—was to placethe beads in a large pan packed with ash or another me-dium to prevent them from sticking together, then theywere reheated while stirring until the edges slumped.This resulted in beads with shapes that are more or lessthe same on both ends. All m-Na-Al beads were treated inthis manner. In contrast, most of the Zanzibar v-Na-Cabeads were rounded in an unusual way: the cut tube seg-ments were placed on a flat surface and then reheated fora short time resulting in beads that are somewhat flat onone end and slightly rounded on the other (Fig. 17:UU229, UU231, and UU063). The effect can be subtleand difficult to see in a photograph since many of thebeads were not reheated for long. Finally, all polychrome,wound, and/or unusually shaped (i.e., segmented) beadsare made of v-Na-Ca glass.

Thus, based on color, bubbles in the glass, unusual endtreatment, and shape, it has been possible to determine

Table 7 A comparison of the composition (in weight percent and ppm)of UU173 with the average of the v-Na 3 beads of the Chibuene series(data from Wood et al. 2012)

Chibuene series (v-Na 3n) UU173

SiO2 66.4 ± 2.0 % 68.03 %

Na2O 15.1 ± 0.8 % 18.24 %

MgO 2.3 ± 0.2 % 1.92 %

AI2O3 3.9 ± 0.4 % 2.48 %

K2O 4.9 ± 0.3 % 3.21 %

CaO 3.7 ± 0.3 % 3.88 %

Cr 20± 3 60

Rb 52± 7 21

Zr 78± 12 79

Cs 2 ± 0.2 n/a

Ba 705 ± 160 148

La 16± 2 4

Nd 13± 2 5

U 3± 1 0.5

Archaeol Anthropol Sci (2017) 9:879–901 895

00.050.1

0.150.2

0.250.3

0.350.4

0.450.5

Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu

v-Na-Ca A

v-Na-Ca B

v-Na-Al

Fig. 16 Average REE pattern ofthe Zanzibar plant ash glassesnormalized to REE in thecontinental Earth’s crust(Wedepohl 1995)

that 14 (46.7 %) of the Fukuchani beads are made of v-Na-Ca glass and 16 (53.3 %) are m-Na-Al glass. AtUnguja Ukuu, only 36 (4.2 %) are made of v-Na-Ca glass,while the remaining 827 (95.6 %) are made from m-Na-Alor v-Na-Al glass (unfortunately, it is not possible to

separate the v-Na-Al beads from the m-Na-Al ones basedsolely on morphology but given that v-Na-Al glass is rareat Unguja Ukuu and on the East Coast in general, it isunlikely that they form more than a small portion of theassemblage).

896 Archaeol Anthropol Sci (2017) 9:879–901

Fig. 17 Unguja Ukuu and Fukuchani tested glass beads. Beads in the top left panel are from Fukuchani, all the rest are from Unguja Ukuu and aregrouped by color. As can be seen in the lowest center panel, UU174 is not a bead but probably a glass weight

Beads and Indian Ocean trade patterns in the firstmillennium CE: East Africa and southern Africacompared

East Africa

Two main types of glass that date to the second half of the firstmillennium CE have been identified in the beads fromUngujaUkuu and Fukuchani. Based on bead distribution in trench 11(see Table 8), which had the deepest stratigraphy (2.92 m) inthe 2011 Unguja Ukuu excavations, it appears possible that m-Na-Al 1 beads—probably made in Sri Lanka or southernIndia—were the first to arrive. As can be seen, the threedeepest, and thus earliest, contexts are made up almost entirelyof m-Na-Al 1 beads. Of the 192 m-Na-Al 1 beads in thosecontexts, 167 are translucent blue-green. Juma (2004: Table7.2.1), in his excavations at Unguja Ukuu, also recorded apreponderance of drawn translucent blue-green beads inPeriod Ia (500 to 750 CE): they accounted for 145 of the178 glass beads recovered. Thus, it appears likely that m-Na-Al 1 beads are the earliest glass beads yet recognized onthe East Coast.

The second type of glass, whose beads probably arrivedsomewhat later than the m-Na-Al 1 beads but ended produc-tion at about the same time, is a plant ash glass (v-Na-Casubtype A) that was made in the Middle East, possibly eastof the Euphrates. The beads, however, may well have beenmade elsewhere since that region is not known for makingdrawn beads and this glass was widely traded in raw and culletform (Lankton and Dussubieux 2006, p. 135, 2013, p. 431;Carboni 2013, pp. 347–348). Regions where the beads couldhave been made include Thailand and Sri Lanka/South India,but other regions, including the Middle East, cannot be ruled

out. However, given that m-Na-Al 1 beads from South Asiawere being traded to Zanzibar, it would not be unreasonable toentertain the possibility that some of the v-Na-Ca beads couldhave been produced in that region or in Thailand, which wasmaking beads from this glass and, in its role as a key link ineast-west commerce at the time, was trading actively with SriLanka and India. The simple monochrome drawn v-Na-Cabeads that were briefly reheated on a flat surface, as found atThung Tuk, Thailand, would be the most likely to fit thisprofile. The more complicated eye beads and striped beadsmay have been made in Thailand, as suggested by Francis(2002, p. 97), or they could have been produced in theMiddle East. In either case, they could have arrived inZanzibar through the South Asian trade links under discussionbecause such beads are not uncommon in South and SoutheastAsia, especially the BTakua Pa^ eye beads which were foundin large numbers at Thung Tuk and were present at Mantai aswell.

A few beads made of v-Na-Ca subtype B glass with ele-vated lime levels were found at both Unguja Ukuu andFukuchani. This glass was also made in the Middle East butpossibly farther to the west. It is not known where the beadswere produced but, apart from the segmented beads, the tech-niques used to produce them are the same as those of the cut orpinched and flat reheated v-Na-Ca subtype A beads, suggest-ing that they were all made in the same region.

Southern Africa

Evidence of Indian Ocean trade to southern Africa begins inthe seventh century CE in the form of beads, made from asubtype of v-Na-Ca glass (v-Na-Ca 3), called the Chibueneseries (Wood et al. 2012). As has been noted (Table 6 andFig. 10), it is a distinct subtype but forms part of the plantash glasses produced in the Middle East. Beads made of thisglass type are rare and seem to have had a short life span.None have been found at any site in East Africa, includingthe Zanzibar sites in this study. In southern Africa, they werefound at Chibuene and a few sites in Botswana as far west asthe Tsodilo Hills (Wilmsen and Denbow 2010; Daggett et al.Glass trade beads at Thabadimasego, Botswana: analyticalresults and some implications, in publication; Denbow et al.2015). Theywere replaced by beads known as the Zhizo seriesmade from a related glass, v-Na-Ca 1 (Robertshaw et al.2010b). As has been shown (Table 6 and Fig. 10), this glassis closely related to Zanzibar subtype A but the morphologicaldifferences between the two sets of beads make it unlikely thatthe beads themselves were manufactured in the same place.Almost all Zhizo beads are simply made by cutting beadlengths from drawn glass tubes; they are not reheated so theirends remain fairly sharp and often rather ragged. They arenever pinched, segmented, or decorated (multicolored)—characteristics that are found in Zanzibar subtype A beads.

Table 8 m-Na-Al 1 andv-Na-Ca beads fromtrench 11 by context

Trench 11 beads

Context m-Na-Al 1 v-Na-Ca

002 1

003 1

004 5 4

005 1

006 2 2

007 2

011 1

012 1 1

014 20 3

016 4 1

017 158 4

018 16

019 18

Totals 229 16

Archaeol Anthropol Sci (2017) 9:879–901 897

A few v-Na-Ca beads that do fit the Zhizo series chemicallyand morphologically have been found in East Africa, but onlyone site has produced more than one and that site is Tumbe onPemba Island (Fleisher and LaViolette 2013) which also pro-duced the cobalt blue bead with white stripes discussed above(section Bv-Na-Ca subtype A^).

East and Southern Africa compared

To Medieval Arab geographers, the eastern African coast com-prised three sections, Bilad al-Barbara, Bilad az-Zanj, and Biladas-Sufala. While the exact boundaries were not fixed, Barbarawas the coast north of Mogadishu, Zanj southwards from thereto the Rovuma River, and Sufala between the Rovuma and theLimpopo (Trimingham 1975, p. 120). The bead evidence fromZanzibar and southern Africa suggests that the Zanj and Sufalacoasts may have been linked to distinct trading systems in thelate first millennium. Historical sources offer some support forthe possibility that the Medieval period saw direct sailing toSufala from the Persian Gulf region that avoided the EastAfrican ports apart from Qanbalu, a port that is thought to havebeen located on the island of Pemba (Trimingham 1975, pp.122, 135; Chittick 1977, p. 192; Hourani 1995, p. 148; Hortonand Middleton 2000, p. 66). Al-Mas’udi, for example, notedthat Sufala was Bthe extreme limit to which the Umanis andSirafis go on the coasts of the sea of the Zanj….the sea of theZanj ends with the land of Sufala and the Waq Waq whichproduced gold and many wonderful things^ (Pellat 1962, III,5). The anonymous tenth-century Authentic Tales of the Seamentions voyages to Sufala in six different tales. One describesa sea journey to Sufala in which the winds and currents drovethe ship onto the Bcoast … despite the captains efforts^(Freeman-Grenville 1981, p. 104). It describes the normal placeBwhere the ships go^ as located 42 zams (approximately 500miles) beyondQanbalu, a location close toMozambique island.Both gold and ivory were key attractions of southern Africa.The Persian scholar Al-Biruni described the actual gold trade,recording that Bat Sufala of the Zanj there is gold of extremeredness^ and noting that the trading practices of the seagoingmerchants involved the local ruler and his elders giving them-selves up as hostages, while Bthe goods that their people desire^were taken inland to trade for gold (Levtzion and Hopkins1981, pp. 58–59; Said 2007, pp. 205–208).

Taken together, these sources provide some evidence thatships from Oman and the Persian Gulf sailed to southernAfrica to trade for ivory and gold, among other goods.Given the continuity in beads involved in that trade duringthe period between the eighth and mid-tenth centuries, alongwith the near absence of those beads at East African sites, itwould not be unreasonable to propose that this trading patternexisted throughout that time span. Al-Mas’udi noted that thesevoyages included a stopover at Qanbalu on their way south(Freeman-Grenville 1962, pp. 14–16). Thus, the observation

that beads found at Tumbe (on Pemba Island) can be linked tothose that are otherwise primarily found in southern Africacould be related to the sailing pattern mentioned by al-Mas’udi. It is tempting to think that the Bgoods the peopledesire^ included glass beads. The trade of Sufala seems tohave been well organized and, it is possible to argue, on asomewhat different basis from that of the ports further northon the Swahili coast.

Bead production

Because Zhizo beads were made of glass that is probably fromthe Iraq/Iran region, were possibly carried to southern Africaon ships fromOman and the Gulf, and are rarely found outsideof southern Africa, it appears likely they were made in thePersian Gulf region or nearby. However, because they weremade from drawn tubes, the artisans who made them wereprobably South Asian since this technology is recognized asbeing South Asian and tube drawing is not known to havebeen practiced in the Iraq/Iran region)2. The absence or rarityof Chibuene and Zhizo series beads outside of southern Africais also significant: had they been produced in South orSoutheast Asia (as is possible with both types of v-Na-Cabeads found in Zanzibar), one would expect that the traderswho supplied those v-Na-Ca beads to the east coast wouldhave carried Zhizo beads as well.

Thus, one could propose that between the eighth and mid-tenth centuries, Zhizo series beads were being made by SouthAsian artisans somewhere in the greater Persian Gulf regionusing local glass and were then carried to southern Africa byships from Oman and the Gulf. Another scenario, however,could be imagined in which South Asian artisans, who livedand worked in the region where the v-Na-Ca subtype A glasswas being made, produced glass tubes in a variety of diame-ters and four basic colors (cobalt blue, yellow, blue-green, andgreen) which were then shipped to widespread destinationswhere they were made into beads based on local tastes andskills. In this case, Zhizo beads would have been cut fromimported tubes at Chibuene, where the skills to reheat themto round the ends were not available or perhaps consumers inthe region preferred unrounded beads. Possible evidence forthis proposal is present in the Chibuene glass assemblagewhich includes a number of useless tube ends, tubes with noperforation and other potential beadmaking debris (Wood etal. 2012). Other sites in Africa that might have worked beadslocally from imported v-Na-Ca glass tubes include Gao inMali (Cissé 2011; Cissé et al. 2013) and Igbo Ukwu in theNiger Delta; both date to around the ninth century. At Gao,tube segments were reheated on a flat surface but for longer

2 For a discussion of artisans moving around the Indian Ocean, seeBellina (2003), Horton (2004), Ray (2004), and Francis (2002), pp. 35–36.

898 Archaeol Anthropol Sci (2017) 9:879–901

periods and/or at higher temperatures than those fromZanzibar, resulting in beads that are more rounded on oneend, flatter on the other, and shinier (MW, personal observa-tion). The Igbo Ukwu beads were ground flat after cutting(MW personal observation), a process that could have easilybeen undertaken locally; in addition, it is rarely seen else-where with these types of beads.

There is no physical evidence that the Zanzibar beads wereworked at Unguja Ukuu or Fukuchani and in any case thevariety of treatments found in those plant ash beads (reheatedon a flat surface, pinched from large tubes, segmented, deco-rated—and two subtypes of glass were involved) suggeststhey were produced either at several different locations or ata sophisticated complex one. In addition, their similarity tobeads found in Thailand and Sri Lanka, which both have am-ple evidence of sophisticated beadmaking, suggests that it ismore likely that they were made in one of those regions ratherthan in Zanzibar. Although the greater Persian Gulf regioncannot be ruled out as a source, beads of these types, apartfrom eye beads, are not common at port sites, such as Siraf,and are absent from Nishapur (although it must be admittedthat glass bead studies in this region are woefully limited).

Africa and its trade with the Indian Ocean

Like other types of material culture, including ceramics anddiverse metal objects, the glass beads that reached Africa orig-inated in a variety of far-flung regions of the Indian Ocean.That many of themwere channeled through theMiddle East issupported by both historical and archaeological evidence, al-though the possibility of some direct trade with South andSoutheast Asia cannot be ruled out. Zanzibar’s glass beadsseem to originate mainly from the latter region, in contrast tothose from southern Africa. This mirrors a key difference inthe imported ceramic assemblages from Unguja Ukuu onZanzibar and Chibuene on the southern African coast –Chinese ceramics are present, albeit in small quantities, atUnguja Ukuu, but are absent from Chibuene (Sinclair 1982;Sinclair et al. 2012, p. 728). Whether this reflects some inputfrom direct trade to Zanzibar as opposed to further south, orwhether it reflects different trading axes from the Middle East,remains to be determined. The possibility that agency on thepart of African consumers may have played some role in thedifferences between the northern and southern sections of theeast coast could be entertained for the East Coast, becausetraders from different regions appear to have been active thereand the diverse types of beads, of different origins, that arefound in assemblages from all periods support this possibility.Choice existed so could be exercised. However, in southernAfrica, agency took a different and very conservative form:only one bead type from one source region was available (or atleast in use) at any period between the seventh and seven-teenth centuries and most of these bead Bseries’^ were

imported for more than a century or two (Wood 2005,2011), being displaced only when the bead type went out ofproduction (i.e., the Zhizo series) or because of significantchanges in trade patterns that involved southern Africa.Portuguese traders learned of this conservatism in the earlysixteenth century when they tried to introduce Europeanbeads—no one would accept them, so the Portuguese (andeventually other European traders) were forced to buy beadsfrom India for the southern African trade (Wood et al. 2009).This pattern continued up to the late seventeenth century whenEuropean beads were finally accepted, partially becauseEuropean beadmakers began to copy the Indian beads.

Conclusion

The glass beads from Unguja Ukuu and Fukuchani open newinsights into Indian Ocean trade to Africa’s eastern seaboard.Two major glass types are present at these Zanzibar sites:mineral-soda (m-Na-Al) and plant ash soda (v-Na-Ca). Mostof the mineral-soda beads tested with LA-ICP-MS belongsecurely to a glass subtype (m-Na-Al 1) that was probablyproduced in Sri Lanka or South India, where the beads wouldlikely have been made as well. This is the first time this glasssubtype has been identified in East Africa or southern Africa.The fact that it was not produced after the tenth century CEconfirms that the trade that brought the beads predates thesecond millennium. Most of the plant ash beads were madeof glass (v-Na-Ca subtype A) produced in the Middle East,possibly east of the Euphrates. A smaller number of relatedbeads (v-Na-Ca subtype B) were made of glass that may havebeen produced farther to the west. Based on method of man-ufacture and the presence of similar beads at other archaeo-logical sites, however, it is possible that many of these beadswere made in South or Southeast Asia. If this were indeed thecase, they may well have arrived in East Africa via the sametrade circuits as the m-Na-Al 1 beads. On the other hand, thebeads found in southern Africa are likely to have come direct-ly from the Persian Gulf region suggesting that during thelatter part of the first millennium CE glass beads traded intoEast Africa and southern Africa arrived via two distinct tradenetworks.

The patterns observed in the Zanzibar bead data, and thecontrasts with southern African datasets, allow us to pro-pose a few possible scenarios for the late first millenniumbead trade on the eastern African coast. The patterns mayreflect the existence of distinct trading axes between theMiddle East and the two sections of the African easternseaboard, rather than a simple extension of the EastAfrican trade down to the southern coast. Or it is possiblethat they reflect the involvement of Zanzibar in a degree ofdirect trade with southern Asia that did not extend to thesouthern African coast.

Archaeol Anthropol Sci (2017) 9:879–901 899

Acknowledgments We would like to sincerely thank Jim Lankton fordata on beads from Thung Tuk and Laure Dussubieux for m-Na-Al dataas well as for comments on parts of the draft. Our paper was improved bycomments from two anonymous referees. Fieldwork was funded by theSealinks Project under a European Research Council Grant (AgreementNo. 206148), awarded to Nicole Boivin.

Open Access This article is distributed under the terms of the CreativeCommons At t r ibut ion 4 .0 In te rna t ional License (h t tp : / /creativecommons.org/licenses/by/4.0/), which permits unrestricted use,distribution, and reproduction in any medium, provided you giveappropriate credit to the original author(s) and the source, provide a linkto the Creative Commons license, and indicate if changes were made.

References

Bellina B (2003) Beads, social change and interaction between India andSouth-east Asia. Antiquity 77:285–297

Boivin N, Crowther A, Helm R, Fuller DQ (2013) East Africa andMadagascar in the Indian Ocean world. J World Prehist 26:213–281

Boivin N, Crowther A, Prendergast M, Fuller DQ (2014) Indian Oceanfood globalisation and Africa. Afr Archaeol Rev 31:547–58131

Bopearachchi O (1999) Sites portuaires et emporia de l’ancien Sri Lanka,nouvelles données archéologiques. Arts Asiatiques 54:5–23

Bopearachchi O (2002) Les relations commerciales et culturelles entre SriLanka et Inde du sud: Nouvelles données archéologiques etépigraphiques Cahier du Cercle d’Études et de Recherches SriLankaises. 4:1–16

Brill RH (1995) Chemical analyses of some glass fragments fromNishapur in the Corning Museum of Glass. In: Kröger J (ed)Nishapur: glass of the Early Islamic Period. The MetropolitanMuseum of Art, New York, pp 211–233

Brill RH (1999) Chemical analyses of early glasses. The CorningMuseum of Glass, New York

Brill RH (2001a) The glassmakers of Firozabad and the glassmakers ofKapadwanj: two pilot video projects. In: Annales du 15e Congres del'Association Internationale pour l'Histoire du Verre. CorningMuseum of Glass, Corning, NY, pp 267–268

Brill RH (2001b) Some thoughts on the chemistry and technology ofIslamic glass. In: Carboni S, Whitehouse D (eds) Glass of theSultans. Yale University Press, New Haven, pp 25–45

Brill RH (2009) Chemical analyses. In: Bass GF, Brill RH, Lledó B,Matthews SD (eds) Serçe Limani: the glass of an eleventh-centuryshipwreck, vol II. Texas A&MUniversity Press, College Station, pp459–496

Carboni S (2013) The Mantai Glass. In: Carswell J, Deraniyagala S,Graham A (eds) Mantai—a city by the sea. Linden Soft Verlag,Aichwald, pp 313–348

Chaisuwan B (2011) Early contacts between India and the AndamanCoast in Thailand from the second century BCE to eleventh centuryCE. In: Manguin PY, Mani A, Wade G (eds) Early interactionsbetween South and Southeast Asia: reflections on cross-cultural ex-change. ISEAS Publishing, Singapore, pp 83–112

Chirikure S, ManyangaM, Pikirayi I, PollardM (2013) New pathways ofsociopolitical complexity in Southern Africa. Afr Archaeol Rev 30:339–366. doi:10.2307/42641836

Chittick N (1977) The East Coast, Madagascar and the Indian Ocean. In:Fage JD, Oliver R (eds) The Cambridge history of Africa, vol 3.Cambridge University Press, Cambridge, pp 183–232

CisséM (2011)Archaeological investigations of early trade and urbanismat Gao Saney (Mali). PhD dissertation, Rice University, Houston

Cissé M, McIntosh SK, Dussubieux L, Fenn T, Gallagher D, ChippsSmith A (2013) Excavations at Gao-Saney: new evidence for settle-ment growth, trade, and interaction on the Niger Bend in the firstmillennium CE. J Afr Archaeol 11:9–37

Crowther A et al (2014) Iron Age agriculture, fishing and trade in theMafia Archipelago, Tanzania: new evidence from Ukunju Cave.Azania 49:21–44

Crowther A, Veall M-A, Boivin N, Horton MC, Kotarba-Morley A,Fuller DQ et al (2015) Use of Zanzibar copal (Hymenaea verrucosaGaertn.) as incense at Unguja Ukuu, Tanzania in the 7–8th centuryCE: chemical insights into trade and Indian Ocean interactions. JArchaeol Sci 53:374–390

Denbow J, Klehm C, Dussubieux L (2015) The glass beads of Kaitshàa:new insights on early Indian Ocean trade into the far interior ofsouthern Africa. Antiquity 89:361–377. doi:10.15184/aqy.2014.50

Duckworth CN, Córdoba De La Llave R, Faber EW, Govantes EdwardsDJ, Henderson J (2015) Electron microprobe analysis of 9th–12thcentury Islamic glass from Cordoba Spain. Archaeometry 57:27–50

Dussubieux L (2001) L’Apport de l’ablation laser couplée à l’ICP-MS àl’étude du verre archéologique de l’Ocean Indien. Universitéd’Orléans

Dussubieux L, Gratuze B (2013) Ancient glass in South Asia. In:Janssens K (ed) Modern methods for analysing archaeological andhistorical glass. Wiley, New Jersey, pp 397–412

Dussubieux L, Kusimba CM (2012) Glass vessels in Sub-Saharan Africa:compositional study of some samples from Kenya. In: Liritzis I,Stevenson C (eds) The dating and provenance of obsidian and an-cient manufactured glasses. University of New Mexico Press, NewMexico, pp 143–156

Dussubieux L, Kusimba CM, Gogte V, Kusimba SB, Gratuze B, Oka R(2008) The trading of ancient glass beads: new analytical data fromSouth Asian and East African soda-alumina glass beads.Archaeometry 50:797–821. doi:10.1111/j.1475-4754.2007.00350.x

Dussubieux L, Robertshaw P, GlascockMD (2009) LA-ICP-MS analysisof African glass beads: laboratory inter-comparison with an empha-sis on the impact of corrosion on data interpretation. Int J MassSpectrom 284:152–161. doi:10.1016/j.ijms.2008.11.003

Dussubieux L, Gratuze B, Blet-Lemarquand M (2010) Mineral soda alu-mina glass: occurrence and meaning. J Archaeol Sci 37:1646–1655.doi:10.1016/j.jas.2010.01.025

Fleisher J, LaViolette A (2013) The early Swahili trade village of Tumbe,Pemba Island, Tanzania, AD 600–950. Antiquity 87:1151–1168

Fleisher J, Wynne-Jones S (2011) Ceramics and the early Swahili:deconstructing the early Tana tradition. Afr Archaeol Rev 28:245–278. doi:10.2307/41486780

Francis P Jr (2002) Asia’s maritime bead trade: 300 B.C. to the present.University of Hawai’i Press, Honolulu

Francis P Jr (2013) The beads. In: Carswell J, Deraniyagala S, Graham A(eds) Mantai—a city by the sea. Linden Soft Verlag, Aichwald, pp349–369

Freeman-Grenville GSP (1962) The East African coast—selected docu-ments from the first to the early 19th century. Clarendon, Oxford

Freeman-Grenville GSP (1981) The book of the wonders of India byCaptain Buzurg ibn Shahriyar of Ramhormuz. East WestPublications, London

Freestone IC (2006) Glass production in Late Antiquity and the EarlyIslamic period: a geochemical perspective. In: Maggetti M,Messiga B (eds) Geomaterials in cultural heritage, vol 257.Geological Society of London Special Publication, London, pp201–216

Freestone IC, Gorin-Rosen Y, HughesMJ (2000) Composition of primaryglass from Israel. In: Nenna M-D (ed) Ateliers primaires etsecondaires de verriers du second millinaire av. J.-C. au Moyen-Age, vol 33. Travaux de la Maison de l’Orient Méditerranéen,Lyon, pp 65–84

900 Archaeol Anthropol Sci (2017) 9:879–901

Freestone IC, Greenwood R, Gorin-Rosen Y (2002) Byzantine and earlyIslamic glassmaking in the Eastern Mediterranean: production anddistribution of primary glass. In: Kordas G (ed) Hyalos-Vitrum-glass. Athens, pp 167–174

Gratuze B, Soulier I, Barrandon JN, Foy D (1992) De l’origine du cobaltdans les verres. Rev Archéom 16:97–108

Helm R, Crowther A, Shipton C, Tengeza A, Fuller DQ, Boivin N (2012)Exploring agriculture, interaction and trade on the eastern Africanlittoral: preliminary results from Kenya. Azania: Archaeol Res Afr47:39–63

Henderson J (2013) Ancient glass: an interdisciplinary exploration.Cambridge University Press, Cambridge

Henderson J, Mcloughlin SD, Mcphail DS (2004) Radical changes inIslamic glass technology: evidence for conservatism and experimen-tation with new glass recipes from early and middle Islamic Raqqa,Syria. Archaeometry 46:439–468

Horton M (2004) Artisans, communities, and commodities: medievalexchanges between northwestern India and East Africa. ArsOrientalis 34:62–80

Horton MC (2015) Zanzibar and Pemba: the archaeology of an IndianOcean archipelago. Ashgate, London

Horton MC, Clark CM (1985) Archaeological survey of Zanzibar.Azania: Archaeol Res Afr 20:167–171

Horton MC, Middleton J (2000) The Swahili. Blackwell, OxfordHourani GF (1995) Arab seafaring; expanded edition. Princeton

University Press, PrincetonJuma A (2004) Unguja Ukuu on Zanzibar: an archaeological study of

early urbanism. Department of Archaeology and Ancient History,Uppsala

Kanungo AK (2004) A database of glass and glass beads in India. ManEnviron 29:42–102

Lankton J, Dussubieux L (2006) Early glass in the Asianmaritime trade: areview and an interpretation of compositional analyses. J Glas Stud48:121–144

Lankton J, Dussubieux L (2013) Early glass in Southeast Asia. In:Janssens K (ed) Modern methods for analysing archaeological andhistorical glass. Wiley, New Jersey, pp 415–443

Levtzion N, Hopkins JFP (1981) Corpus of early Arabic sources forWestAfrican history. Cambridge University Press, Cambridge

Mirti P, Pace M, Negro Ponzi MM, Aceto M (2008) ICP-MS analysis ofglass fragments if Parthian and Sasanian epoch from Seleucia andVeh Ardasir (Central Iraq). Archaeometry 50:429–450

Mirti P, Pace M, Malandrino M, Negro Ponzi MM (2009) Sasanian glassfrom Veh Ardašī. J Archaeol Sci 36:1061–1069

Panighello S, Orsega EF, van Elteren JT, Šelih VS (2012) Analysis ofpolychrome Iron Age glass vessels from Mediterranean I, II and IIIgroups by LA-ICP-MS. J Archaeol Sci 39:2945–2955. doi:10.1016/j.jas.2012.04.043

Pellat C (1962) Mujuj al-Dhahab wa-Ma'adin al-Jauhar of al-Mas'udi.revised edn. Les Prairies d'or, Paris

Ray HP (2004) The beginnings: the artisan and the merchant in earlyGujarat, sixth-eleventh centuries. Ars Orientalis 34(34):39–61

Rehren T, Freestone IC (2015) Ancient glass: from kaleidoscope to crys-tal ball. J Archaeol Sci 56:233–241. doi:10.1016/j.jas.2015.02.021

Robertshaw P, Glascock MD, Wood M, Popelka RS (2003) Chemicalanalysis of ancient African glass beads: a very preliminary report.J Afr Archaeol 1:139–146

Robertshaw P, Rasoarifetra B,WoodM,Melchiorre E, Popelka-Filcoff R,Glascock M (2006) Chemical analysis of glass beads fromMadagascar. J Afr Archaeol 4:91–109

Robertshaw P, Magnavita S, WoodM, Melchiorre E, Popelka-Filcoff RS,GlascockMD (2009) Glass beads from Kissi (Burkina Faso): chem-ical analysis and archaeological interpretation. In: Magnavita S,Koté L, Breunig P, Idé OA (eds) Crossroads: cultural and techno-logical developments in 1st millennium BC / ADWest Africa, vol 2.Journal of African Archaeology Monograph Series. Africa MagnaVerlag, Frankfurt, pp 105–118

Robertshaw P, Benco N, Wood M, Dussubieux L, Melchiorre E, EttahiriA (2010a) Chemical analysis of glass beads from medieval al-Basra(Morocco). Archaeometry 52:355–379. doi:10.1111/j.1475-4754.2009.00482.x

Robertshaw P, Wood M, Melchiorre E, Popelka-Filcoff RS, GlascockMD (2010b) Southern African glass beads: chemistry, glass sourcesand patterns of trade. J Archaeol Sci 37:1898–1912

Said HM (2007) The book most comprehensive in knowledge on pre-cious stones (Kitab-al-Jamahir Fi Ma'rifat Al-Jawahi) by AbuRayhan al-Beruni. Adam, New Dehli

Shortland A, Rogers N, Eremin K (2007) Trace element discriminantsbetween Egyptian and Mesopotamian late Bronze Age glasses. JArchaeol Sci 34:781–789

Sinclair P (1982) Chibuene—an early trading site in southernMozambique. Paideuma 28:149–164. doi:10.2307/41409880

Sinclair PJJ, Ekblom A, Wood M (2012) Trade and society on the south-east African coast in the later first millennium AD: the case ofChibuene. Antiquity 86:723–737

Sode T, Kock J (2001) Traditional raw glass production in northern India:the final stage of an ancient technology. J Glas Stud 43:155–169

Trimingham JS (1975) The Arab geographers and the East African coast.In: Chittick N, Rotberg RI (eds) East Africa and the Orient. Culturalsynthesis in pre-colonial times. Africana Publishing Company, NewYork, pp 115–146

van Elteren JT, Tennent NH, Selih VS (2009) Multi-element quantifica-tion of ancient/historic glasses by laser ablation inductively coupledplasma mass spectrometry using sum normalization calibration.Anal Chim Acta 644:1–9. doi:10.1016/j.aca.2009.04.025

Wedepohl KH (1995) The composition of the continental crust. GeochimCosmochim Acta 59:1217–1232. doi:10.1016/0016-7037(95)00038-2

Wilmsen E, Denbow J (2010) Early villages at Tsodilo: the introductionof livestock, crops, and metalworking. In: Robbins L, Campbell A,Taylor M (eds) Tsodilo, mountain of the gods. Michigan StateUniversity Press, East Lansing, pp 72–81

Wood M (2005) Glass beads and pre-European trade in the Shashe-Limpopo region. University of the Witwatersrand

WoodM (2011)A glass bead sequence for southern Africa from the 8th tothe 16th century AD. J Afr Archaeol 9:67–84

Wood M, Dussubieux L, Wadley L (2009) A cache of ~5000 glass beadsfrom the Sibudu Cave Iron Age occupation. S Afr Humanities 21:239–261

Wood M, Dussubieux L, Robertshaw P (2012) Glass finds fromChibuene, a 6th to 17th century AD port in southern Mozambique.S Afr Archaeol Bull 67:59–74

Archaeol Anthropol Sci (2017) 9:879–901 901